Experimental Physics and
Industrial Control System

Chip Watson <[email protected]> · Wed, 26 Feb 1997 08:49:35 -0500

Maren Purves said:

> Which way is archiving going? 

To which I must reply: many. Jefferson Lab is attempting to
gather support for a new component based archiving system
presented as vaporware at the last EPICS collaboration meeting.
We are working with Bob Dalesio in an attempt to coordinate
his development activities with ours, and have also had
discussions with Matthias Clausen at DESY to integrate his
components.

At the upcoming collaboration meeting in Vancouver, these
developments (from all 3 sites) will be presented in a special
session on archiving.

Implementation of the new architecture is in 3 (or more) phases.
Phase one keeps the existing archiver, supports Bob's new 
archiver, provides an archive server (which hides the multiple file
formats from the clients), and provides 3 clients:

	* XARR, Chris Larrieu's archive data browser
	* StripTool modified to pre-load from archive data,
	  and having the ability to scroll backwards in time
	* simple report tool to produce ASCII files

Soon after, I suspect Matthias (comments?) will be able to
integrate his server and his IDL based browser into this
architecture, and LANL could migrate their java browser to
be a client.

Phase one will be ready by the May meeting.

Phase two replaces AR with a client-server architecture, and
will include the following:

	* archive server (to which data is pushed)
	* channel access client data producer (should this
	  be AR compatible ?)
	* cdev client data producer
	* (optional) IOC resident database client data producer
	* java browser for web access

Phase three includes the following features:

	* correlation viewer/browser (not just time based)
	* data aging / compression
	* ???????? your best idea here ;->

The speed at which this gets done will depend upon how many
people volunteer to contribute to these developments.  Jefferson
Lab will keep 1+ people on the project at least until phase 2 
is complete, including fractions of Matt Bickley and Chris
Larrieu.

Chip

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System